EU AI Act Faces Delays as Member States Struggle to Agree on Final Provisions

The European Union’s landmark AI Act, aimed at regulating high-risk artificial intelligence systems, has faced further delays as member states continue to debate key provisions, according to multiple officials and regulatory documents. The legislation, which would impose strict requirements on AI systems used in areas like healthcare, law enforcement, and critical infrastructure, was originally expected to be finalized by the end of 2024 but now faces an uncertain timeline.

What Are the Key Provisions of the EU AI Act?

The EU AI Act, proposed by the European Commission in 2021, categorizes AI systems based on their risk levels, with strict rules for “high-risk” applications. These include requirements for transparency, data governance, and human oversight. The act also bans certain AI practices, such as real-time facial recognition in public spaces, unless explicitly permitted by law.

According to a draft text reviewed by The Wall Street Journal, the final version of the act includes new safeguards for AI systems used in hiring, credit scoring, and judicial decision-making. These provisions aim to prevent algorithmic bias and ensure accountability, but they have sparked concerns among tech companies about compliance costs.

Why Are Member States Divided?

Disagreements among EU member states have slowed progress on the legislation. France and Germany, two of the bloc’s largest economies, have pushed for more flexibility in regulating AI, arguing that overly strict rules could hinder innovation. Meanwhile, smaller nations like Sweden and the Netherlands have urged stronger protections for consumer rights and privacy.

“The challenge is balancing innovation with the need to protect citizens,” said a spokesperson for the European Parliament’s legal affairs committee. “We’re trying to find a middle ground that works for all 27 member states.”

How Does the EU AI Act Compare to Other Global Regulations?

The EU’s approach contrasts with regulatory frameworks in the United States and China. The U.S. has taken a more sector-by-sector approach, with agencies like the FTC and NIST issuing guidelines for AI use. China, on the other hand, has prioritized state-led AI development, with policies that emphasize national security and economic competitiveness.

A 2023 report by the OECD highlighted the EU’s proactive stance, noting that its regulatory model could serve as a blueprint for other regions. However, the report also warned that overly rigid rules might push tech firms to relocate operations to less regulated markets.

What’s Next for the AI Act?

The European Commission has indicated that negotiations will continue through early 2025, with a final vote expected by mid-year. Meanwhile, advocacy groups are urging lawmakers to prioritize transparency and public input. “This is a critical moment for AI governance,” said a representative from Access Now, a digital rights organization. “The rules set today will shape the future of technology for decades.”

As the debate unfolds, the EU’s ability to reconcile innovation with regulation will be tested. The outcome could set a global precedent for how AI is developed, deployed, and held accountable.